Matches in SemOpenAlex for { <https://semopenalex.org/work/W2952204912> ?p ?o ?g. }
- W2952204912 abstract "The performance of text classification has improved tremendously using intelligently engineered neural-based models, especially those injecting categorical metadata as additional information, e.g., using user/product information for sentiment classification. These information have been used to modify parts of the model (e.g., word embeddings, attention mechanisms) such that results can be customized according to the metadata. We observe that current representation methods for categorical metadata, which are devised for human consumption, are not as effective as claimed in popular classification methods, outperformed even by simple concatenation of categorical features in the final layer of the sentence encoder. We conjecture that categorical features are harder to represent for machine use, as available context only indirectly describes the category, and even such context is often scarce (for tail category). To this end, we propose to use basis vectors to effectively incorporate categorical metadata on various parts of a neural-based model. This additionally decreases the number of parameters dramatically, especially when the number of categorical features is large. Extensive experiments on various datasets with different properties are performed and show that through our method, we can represent categorical metadata more effectively to customize parts of the model, including unexplored ones, and increase the performance of the model greatly." @default.
- W2952204912 created "2019-06-27" @default.
- W2952204912 creator A5002722298 @default.
- W2952204912 creator A5009993083 @default.
- W2952204912 creator A5014589925 @default.
- W2952204912 creator A5024873601 @default.
- W2952204912 creator A5079180840 @default.
- W2952204912 creator A5081115247 @default.
- W2952204912 date "2019-02-13" @default.
- W2952204912 modified "2023-09-26" @default.
- W2952204912 title "Categorical Metadata Representation for Customized Text Classification" @default.
- W2952204912 cites W1614298861 @default.
- W2952204912 cites W1832693441 @default.
- W2952204912 cites W1902237438 @default.
- W2952204912 cites W2000491693 @default.
- W2952204912 cites W2026844232 @default.
- W2952204912 cites W2064675550 @default.
- W2952204912 cites W2070246124 @default.
- W2952204912 cites W2107915309 @default.
- W2952204912 cites W2115733720 @default.
- W2952204912 cites W2118020653 @default.
- W2952204912 cites W2140124448 @default.
- W2952204912 cites W2166706824 @default.
- W2952204912 cites W2250539671 @default.
- W2952204912 cites W2251292973 @default.
- W2952204912 cites W2251849926 @default.
- W2952204912 cites W2358418295 @default.
- W2952204912 cites W2470673105 @default.
- W2952204912 cites W2498028333 @default.
- W2952204912 cites W2550132532 @default.
- W2952204912 cites W2556888587 @default.
- W2952204912 cites W2563010554 @default.
- W2952204912 cites W2572342788 @default.
- W2952204912 cites W2753160622 @default.
- W2952204912 cites W2758755084 @default.
- W2952204912 cites W2773787599 @default.
- W2952204912 cites W2889963505 @default.
- W2952204912 cites W2910453440 @default.
- W2952204912 cites W2962739339 @default.
- W2952204912 cites W2962769776 @default.
- W2952204912 cites W2962902802 @default.
- W2952204912 cites W2964308564 @default.
- W2952204912 cites W2964329882 @default.
- W2952204912 cites W6908809 @default.
- W2952204912 cites W71795751 @default.
- W2952204912 cites W2771844222 @default.
- W2952204912 doi "https://doi.org/10.48550/arxiv.1902.05196" @default.
- W2952204912 hasPublicationYear "2019" @default.
- W2952204912 type Work @default.
- W2952204912 sameAs 2952204912 @default.
- W2952204912 citedByCount "0" @default.
- W2952204912 crossrefType "posted-content" @default.
- W2952204912 hasAuthorship W2952204912A5002722298 @default.
- W2952204912 hasAuthorship W2952204912A5009993083 @default.
- W2952204912 hasAuthorship W2952204912A5014589925 @default.
- W2952204912 hasAuthorship W2952204912A5024873601 @default.
- W2952204912 hasAuthorship W2952204912A5079180840 @default.
- W2952204912 hasAuthorship W2952204912A5081115247 @default.
- W2952204912 hasBestOaLocation W29522049121 @default.
- W2952204912 hasConcept C119857082 @default.
- W2952204912 hasConcept C136764020 @default.
- W2952204912 hasConcept C151730666 @default.
- W2952204912 hasConcept C154945302 @default.
- W2952204912 hasConcept C17744445 @default.
- W2952204912 hasConcept C199539241 @default.
- W2952204912 hasConcept C204321447 @default.
- W2952204912 hasConcept C23123220 @default.
- W2952204912 hasConcept C2776359362 @default.
- W2952204912 hasConcept C2777530160 @default.
- W2952204912 hasConcept C2779343474 @default.
- W2952204912 hasConcept C41008148 @default.
- W2952204912 hasConcept C5274069 @default.
- W2952204912 hasConcept C86803240 @default.
- W2952204912 hasConcept C93518851 @default.
- W2952204912 hasConcept C94625758 @default.
- W2952204912 hasConceptScore W2952204912C119857082 @default.
- W2952204912 hasConceptScore W2952204912C136764020 @default.
- W2952204912 hasConceptScore W2952204912C151730666 @default.
- W2952204912 hasConceptScore W2952204912C154945302 @default.
- W2952204912 hasConceptScore W2952204912C17744445 @default.
- W2952204912 hasConceptScore W2952204912C199539241 @default.
- W2952204912 hasConceptScore W2952204912C204321447 @default.
- W2952204912 hasConceptScore W2952204912C23123220 @default.
- W2952204912 hasConceptScore W2952204912C2776359362 @default.
- W2952204912 hasConceptScore W2952204912C2777530160 @default.
- W2952204912 hasConceptScore W2952204912C2779343474 @default.
- W2952204912 hasConceptScore W2952204912C41008148 @default.
- W2952204912 hasConceptScore W2952204912C5274069 @default.
- W2952204912 hasConceptScore W2952204912C86803240 @default.
- W2952204912 hasConceptScore W2952204912C93518851 @default.
- W2952204912 hasConceptScore W2952204912C94625758 @default.
- W2952204912 hasLocation W29522049121 @default.
- W2952204912 hasOpenAccess W2952204912 @default.
- W2952204912 hasPrimaryLocation W29522049121 @default.
- W2952204912 hasRelatedWork W1519302135 @default.
- W2952204912 hasRelatedWork W1556178152 @default.
- W2952204912 hasRelatedWork W1567338489 @default.
- W2952204912 hasRelatedWork W165252003 @default.
- W2952204912 hasRelatedWork W1965294778 @default.
- W2952204912 hasRelatedWork W1978971213 @default.